The manuscript
The Syriac manuscript, found in the collection of the Academia Sinica in Taipei (Library of Historical Documents of the Ming and Qing Dynasties 297453), is unique in many respects.Footnote 1 It is an early Christian service-book from China, the first part of which contains an early marriage ritual in Syriac that is unparalleled in the Mesopotamian tradition of the Church of the East (the so-called “Nestorian” Church).Footnote 2 The Syriac text has been published recently.Footnote 3 Based on its palaeography and codicology, the manuscript should be dated to the twelfth–thirteenth centuries and was in all likelihood used by Syriac-speaking Christians of the Chinese (Jingjiao) tradition.
Christians first appeared in Tang China in the seventh–eighth centuries as part of the mission of the Church of the East.Footnote 4 While nearly non-existent by the fourteenth century, some Christian communities did survive in Chinese Turkestan, Central Asia, and Southern China (e.g., Guangzhou) in the Mongol period.Footnote 5 These communities used Syriac (Eastern Christian Aramaic) as their liturgical and ceremonial language and Sogdian, as far as we know, as their second, everyday language in both the Sogd area and far to the east.Footnote 6 However, it was Early New Persian (so-called Pārsi-e-Dāri) that gradually became the most widespread, from the decline of the use of Sogdian to the rise of the Mongols.Footnote 7 From the thirteenth–fourteenth centuries onward, Old Turkic became the main Christian language on the eastern shoulder of the northern Silk Road, and New Persian in the west. The borderline between Turkic and Iranian remained quite unstable. Within this borderline, the following text, which contains a marriage ritual written in Syriac and New Persian, as well as later glosses in Old Uyghur, is of great importance.Footnote 8 The secular ritual (or contract) likely formed a part of the wedding ceremony celebrated in Syriac in a community belonging to the Church of the East.Footnote 9 Nevertheless, it is necessary to stress that this ritual is unique and does not correspond to the standard East Syriac marriage service.Footnote 10 Nor does the standard modern service contain any New Persian parts.Footnote 11
The manuscript was first reported by a Canadian missionary, Rev. James M. Menzies, who found himself in Beijing in 1927, alongside his fellows from Anyang, waiting to evacuate by ship to escape the oncoming revolution. On the corner of Pig Market Street (豬市大街) and 4th Eastern Street (東四大街), he spotted eight old photographs of a Syriac manuscript in a street sale. The Chinese vendor had no knowledge of their origin, but Rev. Menzies later learned that the originals had been found in a small storeroom situated on the top of the wall surrounding the Imperial Palace, just over the main southern entrance (午門), which P. Saeki calls the “Horse Gate.” In 1927, the Chinese Historical Museum occupied these rooms on the top of the U-shaped wall. The original Syriac document was discovered by the Society for Sorting Historical Documents from the Ming and Qing Times (明清史料整理會, Ming Qing shiliao zhengli hui). The missionaries handed the photos over to others, with semitologist Dr. W. Taylor publishing them. The second lot of folios was discovered by chance in the library of Academia Sinica in Taibei (Taiwan) by Prof. Nicola di Cosmo, who provided me with the photographic reproduction. Initially, the keepers of the library classified the manuscript as Manchu or Old Mongol. It is likely the manuscript was brought there by the Kuomingtang authorities when they were evacuating museum objects from Beijing after the defeat of Chiang Kai-shek. The structure of the manuscript and the place of the New Persian is as follows (see Fig. 1).
Scheme of folios.
Figure 1 Long description
Arrangement of folios within a quire, using a network of lines and nodes. The nodes are labeled with letters and numbers, indicating the sequence and connections. Black circles represent folios within the quire, while white squares denote old numbers. The lines connecting these nodes are labeled with 'Ra' and 'Vb', indicating the sides of the folios. A red square with the number '2' is highlighted with an arrow, emphasizing its position. The diagram includes a legend explaining the symbols used: black circles for folios within the quire and white squares for old numbers.
The Syriac ritual recorded in the manuscript on f. 2b, although originally dating to about the seventh–eighth centuries, was probably copied in the twelfth century. The fact that it has some Old Uyghur glosses points rather to a pre-Mongol period. Christianized New Persian speakers of Central Asia (tarsa) contributed to the process of shaping the physiognomy of early Syro-Chinese Christianity (“Luminous way” or Jingjiao).Footnote 12 We know already of some Early New Persian texts of that date written in Syriac script, all from the Tarim basin (Turfan, Qočo, etc.). Church and religious texts in Syriac and Sogdian came to be used by local Christians. Later, with the rise of Early New Persian (dāri) in Khorasan and Transoxiana, it came to be used as a kind of lingua franca on the trade ways.Footnote 13 Some Early New Persian texts, while showing some traces of Arabic influence, still used Syriac graphic.Footnote 14 The fragment MIK III/12 from Bulayiq (Turfan area in Xinjiang, China) found by A. von Le Coq is a witness to this tradition. In 1915, Müller analyzed the text, which he described as a Syriac–New Persian psalter. Later, Sundermann added to this text another fragment, SyrHT 153, with the remaining part of the text. Sims-Williams, who corrected the mistakes and misprints of the previous editors, made a new edition of the complete fragment, containing Psalms 131.18–132.113.Footnote 15 This is now catalogued under the number E7 by Sims-Williams.Footnote 16
Furthermore, Sims-Williams reported the existence of two folios from an unpublished pharmacological document: Ms. M 7340 and n. 175. This medical document also testifies to the use of New Persian in Syriac script in the Turfan area.Footnote 17 Thus, it is not a great surprise to meet a New Persian text inside a Syriac-language liturgical book from the eastern part of the Silk Road. As Sims-Williams notes, “Most of the liturgical texts from Bulayiq are in Syriac, but some of them include rubrics in Sogdian Persian.”Footnote 18 Between the Sogdian–Syriac manuscripts there are liturgical specimens as well (e.g., E12, folios from a Syriac baptism liturgy with Sogdian rubrics; E15, Syriac liturgical text of a daily office with a Sogdian rubric).
This looks to be one of the few early-known specimens of New Persian texts written in Syriac script, which can be added to the previous three. Most such texts, as previously mentioned, originated from the eastern part of Central Asia (e.g., Turfan, Qočo, or Toyuk, with Qurutqa less likely). For the sake of comparison, most Uyghur Christian texts from the area originated from Bulayiq, Qurutka, and Qočo.Footnote 19 Although the manuscript was kept in Beijing (北平 Beiping) before being transported to Nanjing and then Taipei, this does not necessarily mean that it was written Beijing or even Kumdān (Chang’an 長安, Xi’an). It is possible that the manuscript was copied from an earlier model connected to Mesopotamia or Iran. The “blessed teacher Ḥnanyā Sop̄ōlōgiōs” mentioned in the manuscript was certainly somebody from western lands.Footnote 20
Now, let us turn to the structure of the Syriac marriage rite from the manuscript, which gives an idea of the setting and internal context of the Persian fragment. The three parts of the wedding rite in the manuscript are as follows: 1. betrothal (parts F–I); 2. marriage formulaic part or contract written in New Persian (the object of the present article); 3. the distribution of ḥnānā and the blessing of rings/ܛܟܣܐ ܕܛܥܝܢܗܬ ܥܳܙܳܩܬܐ ṭaksā d-ṭʕīnūṯ ʕezqāṯā (parts J–N). Liturgically, these are separate parts with different histories. The betrothal (usually məḵuraya) was initially a secular ceremony formalizing the wedding promise, which appeared in the Hellenistic Jewish context and then became well-rooted in the Christian tradition.Footnote 21 However, in the manuscript under examination here, the betrothal is combined with the Persian kabēn, which is the handing of a dowry as part of a marriage contract in Iranian tradition. The marriage itself (burrāḵā or benediction in Syriac) is a different rite that developed from the ancient (Greek and Roman) ceremony of the blessing of the new couple by their parents, which was then remodeled around the sixth century according to eucharistic liturgy. The marriage consisted of several symbolic actions: blessing of the wedding cup, drinking of the blessed wine, giving and exchanging rings, and crowning of the newlyweds. It is perhaps also important to note that the clerical role was initially more that of a witness than an operator of the sacred mystery.
All of this points to the conclusion that the text of the civil part of the ritual looks like an abridged “contract” included in the text of the more protract church service. Provided that the marriage itself took on its ecclesiastical character in Western churches only after the sixth century, it has remained semi-secular in Church of the East communities. In its present form, marriage still preserves traces of being a mixture of secular and sacred elements.
Language and script
Syriac and its influence
The Syriac of the main part is a literary idiom with many liturgical clichés, corresponding for the most part to the Syriac used in Iraq, Iran, and Central Asia. The text is ceremonial, so it quite naturally contains some Syriac technical terms (ḥnānā, masnaftā, etc.). There are some abbreviations and quite a lot of scribal mistakes testifying to the likely degrading knowledge of the liturgical language among Christians of the region. Syriac could have coexisted with Sogdian; later, Sogdian yielded its place to Early New Persian (ENP) and then to Old Turcic (Uyghur). The use of Syriac graphic was quite typical for the Christian variety of ENP.Footnote 22 The connection between the religious and cultural profile and script type was known from Mesopotamia, where garshuni was used by Christians and Arabic script used by Muslims. Quite a number of Sogdian Christian texts were written with the Eastern Syriac alphabet, while all Buddhist and some Manichaean texts used the proper Sogdian script (apart from those that used a special Manichean script).Footnote 23 The fact that Uyghur readers of the later period had difficulties interpreting Syriac (but not Persian) is indicated by several notes in Old Uyghur script and language. The Syriac text and Uyghur marginal notes are the subject of a separate publication.Footnote 24 The New Persian part contains no Uyghur notes. Perhaps Persian, unlike Syriac, retained its presence in Chinese Turkestan and did not need explanatory notes.
The manuscript uses an allographic style of writing (syrography preceding or running parallel to the use of Arabic and Hebrew letters in ENP), which likely points to a Sogdian Christian tradition.Footnote 25 Sims-Williams has given the best description: “The Persian text is written in Syriac script, making use of some non-Syriac letters and conventions.”Footnote 26 In fact, it appears that all Christian ENP texts date from the eighth–twelfth centuries and make use of the Eastern Syriac script (madnḥāyā). As Sims-Williams explains:
some of these letters and conventions are identical to those commonly found in Sogdian texts in Syriac script, in particular the use of sāḏe <ṣ> to represent [c] and [j]; and the new characters <f> (a modified pe) and <x> (a modified kaph) to represent [f] and [x] respectively.Footnote 27 The absence of the Sogdian character <z> (a modified zain) may be a matter of chance, since the sound [z] is rare in Persian and no word in which it would be expected appears in this text. On the other hand, the letters qoph <q>, teth <ṭ>, and ˤe <ʕ> are only used in Syriac and Arabic words and are not used (as in Sogdian) for the sounds [k], [t], and [y]. Instead, [k] is written by means of kaph <k>, [t] by means of taw <t>, and [y] by means of a modified gāmal, here transliterated <g>.Footnote 28
The scribe used a brush that produced less accuracy in letter forms than a calamus. The placing of dots both in Syriac and New Persian is somewhat messy, especially the plural sign (syāmē). More observations on concrete particularities are added in the commentary.
As far as the paleography is concerned, the manuscript is written with a Central Asian madnhaya.Footnote 29 The ductus has some particular traits typical of Nestorian Central Asian: waw is always connected on both sides, alaph ends like the estrangela-type, and the final kaph has a special form.Footnote 30
The tentative profile of the Christian New Persian-speaking community
The community that used the rite was definitely Christian, not Muslim or Manichean. This can be deduced from the following: there is no sign of either Muslim or other (Zoroastrian or Manichaean) features; the Syriac used is quite close to the typical Central Asian variety used in the Church of the East; and in other parts of the manuscript, there are invocations of the Holy Trinity and Christian martyrs (Phebronia, Sergius and Bacchus, and George). E. de la Vaissière specifies that ENP as a written language emerged after 865 AD in Northern Khurasan and thus reflected all the groups that settled along the Silk Road and traveled its path.Footnote 31 Muslims were not present in Turfan until the Timurid period, but Christian churches and monasteries were.Footnote 32 An eighth-century Bukhara dynasty even chose a cross as its symbol.Footnote 33 The findings in Semirechye (in Ak-Beshim/Suyab in particular) attest to the use of Christian Sogdian. The manuscript testifies to the usage of the psalter, Niceo-Constantinopolitan creed, and the hymn “Gloria in excelsis.”Footnote 34 New Persian liturgical documents included Gospel texts and some hymns, including parts of the Hudra (the collection of hymns and anthems of the Church of the East). All of this points to the existence of two groups: urban Christian populations and monasteries located outside cities. The marriage ceremony was intended for the former.
Were the members of this community former Sogdians or Turkic in origin? It is difficult to say. Certainly, ENP was their spoken language at the time the manuscript was copied. However, the community adopted Old Uyghur in the twelfth or thirteenth centuries, evidenced by the marginal notes that provide glosses in Uyghur, such as “here you must say Credo” or “here all the congregation makes a prostration,” which would have been unnecessary if Persian had continued as the language of everyday communication.
It is not clear what this community should be called. The usual appellative term Christians used for themselves in Tang China was Jingjiao (景教, Luminous teaching), but the text was copied outside of the Tang period and even the more precise “Bosi jiao” (Persian teaching) seems out of place in this context. We can consider these Christians instead as tarsa communities on the way to becoming ärkägun/yelikewen (也里可温). The term tarsa originated in Persian (tarsā, Middle Persian [MP] tarsāg, 𐭲𐭥𐭮𐭢) and was adopted by Sogdians as tarsāk. Footnote 35 Tarsa appears on the Christian inscription of the Xi’an stele, rendered Da suo or that-sa (達娑), and seems to also have been used as a technical term for Syriac Christians in Islamic literature in both Persian and Arabic (ترسا).Footnote 36 We can safely assume that the group’s self-designation was tarsākān. Their continued use of Syriac was probably just because it became associated with the initial mission of Alopuen and the blossoming of Eastern Syriac Christianity in China.
An obvious question arises from this discussion: why did these Christians deviate from the main course of liturgical development and come to use a different marriage ritual than the naṣrāyē, or Eastern Syrians of Iraq and Iran? The ritual, although partially preserved, seems to be quite divergent in its rubrics and order from the standard marriage rite of the Church of the East.Footnote 37 There are two possible reasons for this. First, during the period after the Arab conquest, the liturgical reforms performed by patriarch Ishoyahb III were not fully imposed in the eastern parts of the empire. Consequently, a number of local variants may have circulated, perhaps following earlier practices developed in the liturgical tradition of the Chinese Jingjiao Church due to its coexistence with Taoist and Buddhist communities. Second, the marriage rite, which included a special ritual in ENP, could have been a special variant designed for cross-cultural marriages at a time when ENP served as the lingua franca for different ethnic groups.
Kābı̄n as Iranian custom and its place within the Christian milieu
The marriage contract with a part in ENP obviously reflects Iranian legal influence on Central Asian Christians, as it includes the kābīn (MP kābēn) part, a special ritual of the husband handing over the part of his property that constituted the wife’s “share” (sometimes called “bride token”).Footnote 38 The husband forfeited the kābīn if he initiated a divorce. The payment of the kābīn was an important part of Iranian tradition added to the ʿaqd (marriage) ceremony. The ceremony of kābı̄n was also sometimes called mahrieh (dowry) or shīrbahā (bride price).
As previously mentioned, the Middle Persian (MP) term kābēn/CNP kābı̄n (ܟܐܒܝܢ) is used in the text of the Taipei (Beijing) manuscript (l. 6). As far as proper Syriac Christian tradition is concerned, we do not have much evidence of the Persian marriage custom or payment of kābīn in Syriac texts from Mesopotamia. As such, it should be considered a particular custom of areas in which Christians were subject to Persianate authorities.
There is also not much information on how this ceremony developed in the Christian settings of Turkestan. However, from the Persian tradition of concluding marriage in pre-Islamic times, we know that the first part was a marriage proposal (ḵᵛāstagārī) followed by the betrothal (nāmzadī), which began with candy-eating and (later in Islamic ˁadā) the mahrīyeh. This mahr, a marriage portion or donatio propter nuptias, was symbolically represented by sugar but could also be more substantial gifts, including large sums of cash, lands, houses, slaves, household items, or sheep.Footnote 39 The rite was preceded by signing the ʿaqd-nāma (MP paymān), a special betrothal carol, after which the mehrieh of kābı̄n was paid.
A feature that deserves special attention is that the accompanying Syriac rite corresponded to the distributing of water, oil, and ḥnānā, a special mixture of dust from the tomb of a martyr, which appears to be a Christian interpretation of the Iranian cultic element. The use of ḥnānā in the Church of the East was quite widespread (and is well attested in hagiographic sources) but its function in the marriage ritual is not altogether clear.Footnote 40
In the published MP contract, the structure of the Zoroastrian kābēn is much more developed. However, one must admit that a certain affinity between the MP contract and the kabēn in our manuscript shows the common social customs along the Silk Road. These customs were likely acquired by Eastern Syrians in the Sasanian times, when the Persian authorities did not recognize Christian marriage as valid without the conclusion of the kabēn. The formulary of the kabēn in the text is short and succinct, to the point of being symbolical. There are no other signs besides dowry and the legal deposit.
This published text is the first witness of the wedding custom from the area, as all other witnesses are either too short or of a purely liturgical nature. The Tarsā Christians of the post-Tang period had to abandon the idea of being part of Chinese society and adapt to being on the margins of the empire, on the border with Central Asia. This border was open to traders, who had previously been mostly Sogdians but were becoming increasingly Muslim due to the progressing Islamization and partial Arabization of local Persians. The Christians of the area were therefore exposed to Islamic customs, and this likely affected marriage rituals. Later, this community became Turkicized and ultimately adopted Old Uyghur as their everyday language, as evidenced by the marginal notes in manuscripts written in that language. However, in the twelfth century, customs originating from third to seventh-century Zoroastrian Central Asia and Khorasan may still have been practiced.
The contract concerns a deposit of a sum of money (400+1 dinar) as a dowry and 1,559 drachms as a vikālāt (lawful deposit). The interesting thing here is that one dinar is subtracted from the deposit and added to the marriage portion, so the deposited sum is one dinar more. The symbolism of giving and taking is very clear here. What is the reason for using the double dirhām/dinar monetary system? Was there a tradition linking older and newer money, or implying that a dinar (dēnarion) was less than a dirhām? These are topics open to further discussion.
Text
PhotoFootnote 41
Fol. 2b
Main text
ܒܥܕ ܐܙܐܢ ܒܓܘܝ̈ܕ ܣܗ ܒܐܪ …[
ܐܘܠ ܦܕܪܝ ܦܝܣܪܐ ܩܕܘܘܡ ܘܟܝܠ ܕܘݎܬܪܐ
ܬܘܟܝ ݏܠܢ ܢܝ ܒܪ ݏܠܢ ܡܪ ܕܘܭܬܪܝ ݏܠܢ ܪܐ
ݏܠܐܢܗ ܢܐܡ ܐܙܒܗܪܝ ܦܝܣܪ ݏܠܢ ܩܫܐ
ܝܐ ܡܫܡܫܢܐ ܝܐ ܡܗܝܡܢܐ ݏܠܢ ܢܐܡ
ܕܪܡ ܘܝܟܕܝܢܐܪ ܟܐܒܝܢ ܘܗܙܐܪ Footnote 42ܒܨܗܐܪ ܣܕ
ܘܫܫܨܕ ܟܡ ܝܟ ܕܪܡ ܭܐܡ ܗܡܨܘܢܐ ܢܟܝ ܪܣܡ
ܘܐܝܼܢ ܡܐ ܢܡܐ ܕܐܪ ܐܢܣܬ ܒܘܝܟܐܠܬ ܕܪܘܣܬ
Marginal text
ܒܙܢܝ ܦܕܪܝݎܬ
ܓܘܝܕ ܦܕܪܝݎܬܝܡ
ܟܗ̈ܢܐ ܘܡܫܡ̈ܫܢܐ
ܓܘܐܗ ܒܐܫܢܕ
ܡܫܪ ܡܘܒܐܪܟ
ܓܪܕܐܢܕ
Transliteration and transcription
Translation
[Syriac: …to one another. Now:]Footnote 60 After that let say three times/first to the father of a boy and second to the keeper of a girl/«You, that are so-and-so, son of so-and-so, [give] to the daughter of so-and-so/by the name so-and-so on account of the bridegroom, the priest/or deacon or believer by the name so-and-so,/ four hundred dirhāms and one dinar as the marriage portion and thousand/and six hundred minus one dirhāms as the law-deposit (wikālat),/and this document is our expression as the right proxy.
[Margin:] According to the customFootnote 61 father of the bridegroom says to the father of the bride: “It is accepted,/should the priests and deacon(s)/be witnesses!”/And let him the blessed chamberFootnote 62/prepareFootnote 63».
Immediately after the text of the marriage contract in the manuscript (f. 6v) follows the ṭaksā, i.e., the order of the betrothal:
ܬܘܒ ܛܟܣܐ ܕܡܟܘܪܝܐ ܒܝܘܡܐ ܟܠܬܐ ܠܚܬܢܐ. ܒܬܪ ܕܡܬܡܠܠܐ ܡܠܬܐ ܥܠ ܡܟܝܪܘܬܐ ܡܝܬܝܢ ܡܨܢܦܬܐ ܘܥܙܩ̈ܬܐ ܕܚܬܢܐ ܘܕܟܠܬܐ ܘܫܟܪ ܘܣܝܡܝܢ ܩܕܡ ܨܠܝܒܐ ܘܐܘܢܓܠܝܘܢ
(Next, the rite of betrothal, on the day when the bride is betrothed to the bridegroom. After the word(s) concerning the betrothal are spoken, they bring a crown and the rings of the bridegroom and the bride, and the sugar: and they place (them) before the cross and Gospel).
Grammatical and lexical commentary
In these texts, we are dealing with a Christian variety of ENP with some distinctive features. Such features could have originated in local dialects, as Lazard noted: “its [i.e., ENP] local variants in the different regions where it took root were influenced to a greater or lesser degree by the dialects which it replaced.”Footnote 64 Alternatively, these features may have originated in other languages, such as Sogdian or Arabic. Grammatically, the text seems to use a language closer to the northeastern type of ENP (according to Lazard’s classification) than the southern or western varieties, and there are neither remnants of MP orthography nor Parthian features.Footnote 65 If one accepts this, I suggest a date between the end of the end of ninth century and the end of the twelfth century based on linguistic criteria: the terminus post quem is the end of the ninth century and the post quem non is the end of the twelfth century.
Some grammatical features of the ENP are worth mentioning. The function of the suffix -ra is still a marker of an indirect object (pisar-rā, a boy) as in Classical New Persian (CNP) and some modern Persian varieties of Afghanistan and Central Asia.Footnote 66 Ki (probably with a short vowel), a multifunctional particle typical of ENP, is used (tū kī̆ falān-ē, thou who art so-and-so). The preposition az is used in the sense “from.” Also worth mentioning is the form bāšand (bāš- is derived from the MP sg. imperative bāš) + 3 pl. -and.
Phonetically there is not much to tell. The obvious difference between arabographic usage and the syrography is that the internal <a> and <ā> are written with or without alaph, like in Syriac, and pronounced more like ā than the Persian å. The use of ܗ and ܝ [ē/ī] for the ezāfe is irregular, it can be omitted (like in the arabography) from the position where it is grammatically necessary, such as in pysr pln = *pisar-i falan (l. 4), bwyklʾt drwst = *be vikalāt-i durost (l. 8), and bzny pdr = ba zan-ī piδar (l. 9). The final consonant of the word to which ezāfe connects could be perceived as geminate: pln ny (l. 3). The shift u>i in the first syllable is yet present, so pysr = pisar, probably already not pusar (Av. puθra). Less evident is pdr, which could be dialectal “padar” (as in Tajik) but the way it is spelled suggests a traditional vocalization.
Lexical features betray a mixture of old Syriacisms and newer Arabisms.Footnote 67 This could be explained as just a common feature of mixed communities on the move or, for instance, ecclesiastical terms such as “priest,” “deacon,” or “layman” that are borrowings from Syriac (məšamšānā, kahnā, məhaymānā). It remains unclear whether terms of Middle Persian origin are also used: the contract is called kābı̄n and the expression ba-zan, although Iranian, seems to be related to the Syriac ܒܙܢܐ, “according to the custom,” as well.Footnote 68 The money is called “dirhams” (drm) (a shortened Arabic درهام from Greek δραχμή) and “dinars” (dynʾr), from the name of the currency used by Silk Road traders.Footnote 69 It combined old dinar-based money used in the Khalifate and the newer dirham-based one current in Central Asia. Central Asian dirhams were coined in Bactria as well as in the Semirechye, so these could well have been of local provenance. In the transcription, the Arabic ā is not used, as the pronunciation betrays elements circulating already in Central Asia: h fell out, so perhaps ā was also no longer.
Clearly, the regular Arabic lexica penetrated the ENP, hence some words of Arabic origin: awwal, أول (l. 2); mubārak, مبارك (l. 13); and baʿd, بعد (l. 1). Yet another Arabism is expressed with two words from the same root: vakı̄l (a legal representative, a proxy) and vikālat (agency, power of a lawyer). The first corresponds to the Arabic wākil, واکل (with a different morphological model, whereas FAʕĪL is more common in NP).
Falān/folān (lines 3–5) needs special attention. In the Middle Persian (or Pahlavi) marriage rite text published by Jamasp-Asana and analyzed by Maria Macuch, which reflects an early medieval Mazdaic marriage custom, all the name placeholders are expressed through wahmān, “so-and-so,” which corresponds to the later term falān, of Aramaic origin (Early Aramaic: pe-la-nu but later palān, Syriac ܦܠܳܢ also has zkāp̄ā over the last syllable) and used precisely in our text.Footnote 70 The phonetic variant ݏܠܢ with the Sogdian letter ݏ (f) instead of Syriac p shows that this was still perceived to be a foreign word and pronounced with f and a. Footnote 71 In CNP, it is usually fulān, فلان, from the Classical Arabic falān. Footnote 72 The Syriac ݏܠܢ ܒܪ ݏܠܢ seems to render the formula wahmān ī wahmānān, usually used in the kābēn-contract, so I opted for the ḥarakat-e zabar.
Lines 9–10 use the terms xat/xatı̄m (pdry xt, piḏar-i-xat; pdr xtym, piḏar-i xatim), which seem not very clear. Some scholars have compared xat/xatīm to the Arabic خَتَمَ (“the seal” and later “ring”), thus translating the words as “sealed.”Footnote 73 However, the very opposition xat-xatı̄m makes this translation less plausible. Another hypothetical reading is based on the two phonemes t and ṭ intermingling, thus “handwriting” or “script” (خط), which also do not seem likely translations to me. The word xat could also be a result of a very long and entangled story, being in the ENP a possible borrowing from Sogdian γwtʾynh/xwatēn, xutēn, “queen, noble lady,” in an unattested shortened form. As to the absence in the CNP of khā̆t, one could point to the MP xwadāy and the CP خدای (xodây), certainly not in the sense of “Lord, God” but instead “a nobleman, sire.” It seems likely that it is an abbreviated form of the Syriac ḥaṯnā (related to Akkad. ḫatanu, cf. Arabic ختن), by way of popular etymology connecting it to the Sogdian word for “groom/maiden.”Footnote 74 The Persian analogue to the Syriac kalləṯā (bride) in our text is xtym, which could have been formed from ḥaṯnā by addition of the Turkic 1 sg. possessive suffix -(X)m, which often forms feminine titles.Footnote 75 Certainly, it suggests a Turkic influence on this variety of ENP, which has no parallels to my knowledge.Footnote 76
In sum, the New Persian idiom in the manuscript from Taibei is an early variety of NP, interesting primarily because of its date and use of the Syriac script. Lexically, we see a mixture of the old borrowings from Syriac and new Arabic words that came from Central Asia. Phonetically, there is a distinction of p and f, probably also a spirantization of d (into δ) in an inter-vowel position, as was the case in northeastern dialects.
Value and importance of the text
Historically, this text brings to light an interesting aspect of Christian life in early Medieval Turkestan, in which local Christians, regardless of origin (Iranian, Turkic, Mongolian, or Chinese), used Syriac for liturgical purposes until the rise of the Mongol Empire. The community that used this “service-book” and celebrated the marriage rite accordingly was most likely a mixed one, potentially consisting of merchants and other people on the move who circulated between Iran, Central Asia, and China’s western provinces. Many such people were likely of Iranian origin, possibly of Sogdian descent, but Turkic elements are not excluded. They belonged to the Eastern metropolitan dioceses of the Church of the East (Chach?), which used rites and service-books that differed from those used in the provinces of Iraq or Iran. In particular, the rite prescribed the use of hnana (a mixture of water, oil, and dust from the tomb of a martyr) as a symbolic gift for newlyweds. The community probably continued to function until the late Yuan period, as suggested by the appearance of Old Uyghur glosses. Old Uyghur continued to be used in the region until the fourteenth century, by which point the manuscript had already reached Beijing (at that time, probably already Khan-balık).
This text should be considered an important lens into the history of the Persian language, categorized as an early syrographic NP texts. ENP, used in the manuscript in a portion with the kabēn, replaced Sogdian in everyday usage. The language shows some Arabic borrowings, which makes an eleventh century date quite plausible.Footnote 77 Its phonetics and grammar point to the Syriac religious substratum and an Arabic superstratum. The profile of the ENP corresponds to the northeastern (Transoxanian?) variety.
The text published here should be classified as a civil part of the religious ceremony, a kind of contract (handing over of the dowry), included in a larger wedding ritual performed in Syriac. The qualification of the fragment as such is given in the very text as kābīn.
Acknowledgements
I am profoundly grateful to Dr. Ilya Yakubovich (Philipps Universität Marburg, Institut für Klassische Sprachen und Literaturen, Germany) for his valuable help sharing readings and his immense expertise. Thanks also go to Dr. Sebastian Brock (Oxford), for advising on the Syriac text, and Dr. Nicholas Sims-Williams (SOAS, London) and Dr. Paola Orsatti for their encouragement and suggestions at the initial stage of the research. I thank my colleague Dr. Timur Koraev for our discussion on Central Asian Persian language varieties.
Appendix 1. Syriac graphic used in the manuscript and parallels
